[Neuroscience] Re: Series Resistance compensation, what the hell?
(by r_s_norman from _comcast.net)
Tue Apr 3 20:46:50 EST 2007
On 3 Apr 2007 18:30:30 -0700, "Bill" <connelly.bill from gmail.com> wrote:
>I though I understood Rs compensation, evidendtly I have no idea.
>I thought you figured out Rs, say 10MOhm, the command voltage is 100mV
>the amplifier wants to pass say 1nA, it knows it would generate 10mV
>across Rs, and hence when it thought the cell was at 100mV, it was
>actually at 90mV. Hence compensation multiplies passed current by a
>proportional to Rs, bringing the voltage behind the seal to a value
>above the command voltage, but getting the cell to the correct value
>But I know realize that that logic is flawed, as the voltage drop
>across Rs, is proportional to Rs/Rs + Rin, i.e. a voltage divider.
>So how does the amp know how much voltage is lost across Rs, when that
>is proportional to Rin, a value which changes all the time (synaptic
>currents, voltage gated currents etc)?
The voltage lost across Rs from an external voltage source is
determined by the voltage divider effect. However the portion of the
input voltage caused by externally forced electrode current is simply
given by ohm's, E = IRs. That portion of the voltage is what gets
>Furthermore, reading the axon manual, I do not understand the
>"The amount of compensation achievable is limited by two
>considerations. First, as the compensation level (a) approaches 100%,
>the increase in the command potential hyperbolically approaches
>infinity. For example, at 90% compensation, the command potential is
>transiently increased by a factor of ten (Vcmd/(1 - a)). Thus at large
>compensation levels the electronic circuits approach saturation"
>Transiently? I thought the current was always increased by a factor
>proportional to the compensation %? And why would 100% need infinite
>current to pass, how would this overcome the finite Rs?
I assume the "transient" nature of the command potential increase is
caused simply by the transient nature of the current pulse being
injected. At 90% compensation, 90% of the voltage change is caused by
the series resistance and only 10% by the system under test.
Therefore, the command voltage has to be increased 10-fold so that the
10% the system under test sees is the correct value. At 100%
compensation, essentially all the voltage change is spurious, caused
by the series resistance and essentially zero is caused by the actual
system you are looking at. That is, 100% compensation is caused by
the series resistance being infinitely greater than the cell
resistance. The percent compensation is essentially the voltage
divider you mentioned previously.
More information about the Neur-sci